Investment attraction has been challenging in 2025 with estimates suggesting that project numbers may fall by around 17% by the end of the year. Investors remain cautious, driven by a complex and interconnected mix policy uncertainty, trade tensions, rising protectionism, and sluggish demand.
Yet behind the headline numbers, there are encouraging signals within this year’s FDI landscape. In particular, three technologies stand out as ones to watch in 2026.
Putting AI to work
Unsurprisingly. AI has dominated investment in 2025 for both the number of projects and year-on-year growth. By the end of the year, we could see over 1,000 FDI projects in AI compared to 590 in 2024.
AI investment remains concentrated in two areas. Software which accounts for 65% of projects but only 5% of capital investment, and hardware (eg semiconductors and data centres) which accounts for 21% of projects but 92% of capital investment.
The remaining 14% of investment projects account for only 3% of capital investment, yet this slice may signal the future direction of AI-driven FDI. Venture capital trends are an early signal of FDI activity and in 2025 there was a of diversification of venture capital funding into AI applications beyond pure technology plays. Sectors such as energy, manufacturing, mobility, healthcare, and pharmaceuticals are emerging as attractive fields for AI integration, creating fertile ground for future cross-border investment.
While these projects may not match the capital intensity of AI hardware, supporting them in 2026 could be a smart strategy for investment promotion agencies aiming to showcase how AI strengthens their region’s existing industrial capabilities.
Supporting a world beyond text
Linked to AI has been the continued growth of data centre FDI in 2025, with 280 projects in the first ten months of the year set to outperform figures from 2024 (295 projects). A talking point in 2025 has been the risk of an investment bubble in AI as companies spend vast sums on data centres to train AI models that could eventually run out data. But this argument forgets that there is a universe beyond using text to train AI, and developers are just beginning to explore it.
The next frontier is “world model” data, where data centres will be used to train AI models to understand the dynamics of the real world, including physics and spatial properties. This will drive innovation in robotics, driverless cars and other autonomous systems, and locations that can offer real-world testing of such models will be well placed to attract future investment.
Military technology
While AI is the largest technology area for FDI in 2025, the fastest growing is military technology. Driven by government investment, FDI in military technologies are expected to almost double from 93 projects in 2024 to 143 in 2025, with still two months of data to collect.
This surge reflects two major trends.
Firstly, uncertainty is a global phenomenon meaning investors have a large market to expand into – the top destinations for investment cover all parts of the globe the US, UK, Ukraine, India, Australia, and the UAE.
Secondly, military technology is about more than tanks and bombs as communications systems, deep-tech, and dual-use technologies are all now part of a growing sector. This can be demonstrated by the fact that Europe’s most valuable defence startup, Helsing, is primarily an AI company.
With opportunities in traditional manufacturing and high-tech dual use technologies, the defence sector offers a diverse set of strategic avenues for investment promotion agencies.
Play to your strengths in 2026
Understanding which of these technologies your city or region should promote depends on matching your existing strengths with future trends. OCO Global are experts at combining location mapping with horizon scanning to build future-orientated strategies and compelling narratives for investment.
Contact us to learn more.
Note on sources: all data on FDI projects and investment is from FT Locations: fDiMarkets database. Data was available up to October 2025 and full-year estimates are based on OCO Global analysis of the data.
